LAUSR

Background Besides providing reassurance about cardiovascular (CV) safety of newer diabetes drugs, cardiovascular outcome trials (CVOTs) have also shown encouraging benefits on some CV endpoints. The contribution of the better glycemic control in the reduction of major cardiovascular events (MACE) remains an open question. The aim of this study is to evaluate the associations between the reduction of HbA1c and risk of MACE, MACE components, hospitalization for heart failure (HF) and all-cause death in CVOTs. Methods An electronic search up to July 2021 was conducted to determine eligible trials. Systematic review identified eighteen CVOTs reporting prespecified CV outcomes. Pooled summary estimates and 95% confidence intervals (CI) were calculated according to the random effects model using the Paule-Mandel method; restricted maximum likelihood estimators were used to estimate model parameters in the metaregression. Results The eighteen CVOTs evaluated 161,156 patients and included four trials with dipeptidyl-peptidase-4 inhibitors (DPP-4i), eight trials with glucagon-like peptide-1 receptor agonists (GLP-1RA) and six trials with sodium-glucose cotransporter-2 inhibitors (SGLT-2i). Random-effects model meta-analysis showed an association between treatment and risk of MACE (hazard ratio [HR] 0.90; 95% CI 0.86, 0.94, P < 0.001), with significant heterogeneity between studies (I 2  = 45.2%, Q statistic P = 0.040). In meta-regression, there was an association between the reduction in HbA1c at the end of the trial and the HR reduction for MACE (beta =  − 0.298, P = 0.007), with significant heterogeneity (I 2  = 40%, Q statistic P = 0.04); this association was totally driven by the risk reduction of non-fatal stroke, which explained 100% of between-study variance (beta =  − 0.531, R 2  = 100%), without heterogeneity (I 2  = 24%, Q statistic P = 0.206). There was no association between the reduction in HbA1c and the HR for heart failure or all-cause death. Conclusions The reduction of HbA1c in eighteen CVOTs was significantly associated with reduction of non-fatal stroke, explaining all (R 2  = 100%) of the between-study variance. While the contribution of glucose lowering in some CV benefits of newer agents does not influence their indications for the patient with type 2 diabetes, it may hopefully facilitate their use.